Searching for Celiac Disease Screening-detected celiac disease in an HLA-genotyped birth cohort

(1)

LUND UNIVERSITY

Searching for Celiac Disease Screening-detected celiac disease in an HLA-genotyped

birth cohort

Björck, Sara

2015

Link to publication

Citation for published version (APA):

Björck, S. (2015). Searching for Celiac Disease Screening-detected celiac disease in an HLA-genotyped birth cohort. Diabetes and Celiac Disease Unit.

Total number of authors: 1

General rights

Unless other specific re-use rights are stated the following general rights apply:

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.

• You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal

Read more about Creative commons licenses: https://creativecommons.org/licenses/ Take down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

(2)

Searching for Celiac Disease

Screening-detected celiac disease in an

HLA-genotyped birth cohort

Sara Björck

DOCTORAL DISSERTATION

by due permission of the Faculty of Medicine, Lund University, Sweden. To be defended October 23, 2015 9:00 a.m. in the CRC Aula (CRC

93-10-002), Clinical Research Center, Jan Waldenströms gata 35, Malmö.

Faculty opponent

Professor Katri Kaukinen, MD, PhD

(3)

Organization

LUND UNIVERSITY

Document name

DOCTORAL DISSERTATION Department of Clinical Sciences, Malmö _{Date of issue October 23}rd_{, 2015}

Author Sara Björck Sponsoring organization Title and subtitle

Searching for Celiac Disease. Screening-detected celiac disease in an HLA-genotyped birth cohort

Abstract

Objectives: Celiac disease is a common immune mediated enteropathy strongly associated with HLA-DQB1*02

(DQ2), *0302 (DQ8), or both and the presence of tissue transglutaminase autoantibodies (tTGA). Prevalence studies have revealed that most affected individuals go undetected because of subclinical signs or being asymptomatic rendering screening a method for identification. However, less is known about subclinical manifestations of screening-detected celiac disease during childhood and if these motivate identification and treatment. The overall aim of the present research was to identify children with screening-detected celiac disease in an HLA-genotyped birth cohort and to study systemic cytokines and bone mineral density (BMD) in these children.

Methods: Children were HLA-genotyped at birth and offered screening at three and nine years of age by

detection of tTGA in plasma using radioligand binding assays. Children repeatedly positive for tTGA underwent intestinal biopsy to confirm diagnosis of celiac disease. The cytokines IFN-γ, IL-1β, IL-2, IL-4, IL-5, IL-8, IL-10, IL-12p70, IL-13, and TNF-α were analysed at time of diagnosis and after treatment with a gluten-free diet and compared with matched controls. At nine years of age, children with screening-detected celiac disease were examined by dual X-ray absorptiometry for analysis of BMD and for serum 25(OH) vitamin D3 and plasma parathyroid hormone (PTH) and compared to matched controls.

Results: Screening-detected celiac disease was found in 3.5% (56/1618) of three year old children having

HLA-risk alleles compared with none (0/1815) among children not having these HLA-risk alleles (p<0.001) (Paper I). A follow-up screening at nine years of age identified an additional 3.8% (72/1907) with celiac disease in the HLA-risk group compared with none (0/2167) in the control group (p<0.001) (Paper II). Three-year old children with screening-detected celiac disease had systemically elevated pro-inflammatory cytokines of both TH1 and TH2 pattern compared to controls of which most were down-regulated after starting a gluten-free diet (Paper III). At nine years of age, children with screening-detected celiac disease had lower BMD, lower levels of vitamin D but higher PTH levels compared with matched controls. In contrast, children on a gluten-free diet did not differ from their matched controls (Paper IV).

Conclusions:Screening-detected celiac disease is only found among children at genetic risk but repeated testing during childhood is necessary to detect new patients. HLA-genotyping could therefore be used to select large populations to be screened for celiac disease. Children with screening-detected celiac disease have systemically elevated pro-inflammatory cytokines and low BMD but normal values on a gluten-free diet, indicating that children with screening-detected celiac disease could benefit from early identification and treatment.

Key words Bone mineral density, Celiac disease, Children, Cytokines, DQ2, DQ8, DXA, HLA, Screening,

tissue transglutaminase antibody

Classification system and/or index terms (if any)

Supplementary bibliographical information Language English ISSN and key title 1652-8220

Lund University, Faculty of Medicine Doctoral Dissertation Series 2015:106

ISBN

978-91-7619-185-9

Recipient’s notes Number of pages Price Security classification

I, the undersigned, being the copyright owner of the abstract of the above-mentioned dissertation, hereby grant to all reference sources permission to publish and disseminate the abstract of the above-mentioned dissertation.

(4)

Searching for Celiac Disease

Screening-detected celiac disease in an

HLA-genotyped birth cohort

(5)

Department of Clinical Sciences, Malmö Lund University, Faculty of Medicine Doctoral Dissertation Series 2015:106 ISBN 978-91-7619-185-9

ISSN 1652-8220

Printed in Sweden by Media-Tryck, Lund University Lund 2015

(6)

To all children having celiac disease;

known and unknown.

(7)

(8)

associated with HLA-DQB1*02 (DQ2), *0302 (DQ8), or both and the presence of tissue transglutaminase autoantibodies (tTGA). Prevalence studies have revealed that most affected individuals go undetected because of subclinical signs or being asymptomatic rendering screening a method for identification. However, less is known about subclinical manifestations of screening-detected celiac disease during childhood and if these motivate identification and treatment. The overall aim of the present research was to identify children with screening-detected celiac disease in an HLA-genotyped birth cohort and to study systemic cytokines and bone mineral density (BMD) in these children.

Methods: Children were HLA-genotyped at birth and offered screening at three

and nine years of age by detection of tTGA in plasma using radioligand binding assays. Children repeatedly positive for tTGA underwent intestinal biopsy to confirm diagnosis of celiac disease. The cytokines IFN-γ, IL-1β, IL-2, IL-4, IL-5, IL-8, IL-10, IL-12p70, IL-13, and TNF-α were analysed at time of diagnosis and after treatment with a gluten-free diet and compared with matched controls. At nine years of age, children with screening-detected celiac disease were examined by dual X-ray absorptiometry for analysis of BMD and for serum 25(OH) vitamin D3 and plasma parathyroid hormone (PTH) and compared to matched controls.

Results: Screening-detected celiac disease was found in 3.5% (56/1618) of three

year old children having HLA-risk alleles compared with none (0/1815) among children not having these risk alleles (p<0.001) (Paper I). A follow-up screening at nine years of age identified an additional 3.8% (72/1907) with celiac disease in the HLA-risk group compared with none (0/2167) in the control group (p<0.001) (Paper II). Three-year old children with screening-detected celiac disease had systemically elevated pro-inflammatory cytokines of both TH1 and TH2 pattern compared to controls of which most were down-regulated after starting a gluten-free diet (Paper III). At nine years of age, children with screening-detected celiac disease had lower BMD, lower levels of vitamin D but higher PTH levels compared with matched controls. In contrast, children on a gluten-free diet did not differ from their matched controls (Paper IV).

Conclusions: Screening-detected celiac disease is only found among children at

(15)

patients. HLA-genotyping could therefore be used to select large populations to be screened for celiac disease. Children with screening-detected celiac disease have systemically elevated pro-inflammatory cytokines and low BMD but normal values on a gluten-free diet, indicating that children with screening-detected celiac disease could benefit from early identification and treatment.

(16)

Background

Introduction

Early in my career as a medical student, and later on as a pediatrician, I learnt that celiac disease is a common immune-mediated enteropathy induced by dietary gluten ingestion in genetically predisposed individuals occurring throughout life. The classical symptoms of celiac disease in children are diarrhoea, abdominal distension and failure to thrive. A gluten-free diet, defined as a diet free from wheat, rye and barley, results in a recovery of the intestinal mucosa in the majority of affected individuals, but reintroduction of gluten into the diet results in relapse of the disease even after years of treatment.

During my first summer as a resident physician at the Children´s Hospital in Malmö, Sweden, I met a 3-year old girl approximately 3 years of age with chronic diarrhoea and tiredness. She was pale, had thin arms and legs and a large belly. I especially remember that her mother had to help her to get up from the examination table because she was weak and her belly prevented her from getting up. This was the first patient with classical appearance of celiac disease that I investigated - and the last. Since then, I have treated hundreds of children with celiac disease with a myriad of symptoms, but also children that are completely free from symptoms. This thesis is my contribution to the knowledge about celiac disease in children.

Pathogenesis

To date there are two known prerequisites for the development of celiac disease: having a genetic susceptibility for the disease and being exposed to gluten [1]. Since only a minority of genetically predisposed individuals that ingest gluten develop celiac disease, there are most likely environmental factors involved that may trigger the disease leading to the immune response to gluten [2] (Figure 1).

(17)

Figure 1. Schematic view of the major factors involved in the pathogenesis of celiac disease. The

genetic susceptibility consists mostly of the HLA-DQ2 or DQ8 haplotypes and dietary gluten is the major exposure. In addition, there has to be environmental exposures which in combination elicit the typical immune response in the intestinal mucosa leading to production of autoantibodies such as tissue transglutaminase autoantibodies and the histological characteristics comprising of intraepithelial lymphocytosis and villous atrophy. Recreated from Green et al. J Allergy Clin Immunol, 2015. 135(5): 1099-1106 and published with permission from the publisher.

In celiac disease, two main antigens have been identified; gluten, the exogenous antigen of which is necessary for celiac disease to develop, respective tissue transglutaminase type 2 (tTG), the main autoantigen identified in 1997 to which autoantibodies are produced [3]. In contrast to other autoimmune diseases, the disease process can be reversed by eliminating gluten from the diet resulting in a reduction of tTG autoantibodies (tTGA) [4, 5].

Gluten

Wheat protein is divided into 4 categories based on their solubility in different solvents: albumins (soluble in water), globulins (soluble in dilute salt solutions), gliadins (soluble in aqueous alcohol) and glutenins (soluble in dilute alkali or acid) [6]. Gluten is often used as an aggregate name for the prolamin storage proteins of wheat (gliadins and glutenins), rye (hordeins) and barley (secalines).These proteins are after ingestion first degraded into large fragments by pepsin in the stomach, but due to the richness in the amino acids proline and glutamine the proteins are resistible to degradation by intestinal intraluminal and brush border endopeptidases and are only partially degraded into gliadin peptides before reaching the mucosa of the small intestine [7]. The spacing between proline and glutamine also plays a significant role in the deamidation by tTG [8]. On the

EPITHELIUM Innate respons LAMINA PROPRIA Adaptive respons GLUTEN GENETIC FACTORS HLA DQ2/8 Autoantibodies Intraepithelial lymphocytosis + Villous atrophy ENVIRONMENTAL FACTORS

(18)

contrary, the prolamine proteins in oats, called avenins, contain much less amount of proline, which could explain why the majority of celiac disease individuals tolerate pure oats [9].

Genetics

The observation that celiac disease is more common in relatives to celiac disease individuals and the high concordance rate between monozygotic twins indicate a strong genetic influence on disease risk [10]. The first observed genetic association was to certain alleles in the human leukocyte antigen (HLA) class II region on chromosome 6 [11]. This region codes for the major histocompatibility complex (MHC) class II molecules, which function as antigen receptors on antigen presenting cells (APC). In celiac disease patients, gliadin peptides show high affinity to certain HLA heterodimers leading to a subsequent activation of T cells [12, 13].

The HLA heterodimer comprises of two ligands; the α-ligand and the β-ligand. The genes coding for these ligands reside in the HLA-DQ region in close linkage disequilibrium with the adjacent DR-region. Approximately 90-95% of celiac disease individuals carry the HLA-DQA1*05 and DQB1*02 alleles (coding for the molecule DQ2.5) [14] although carrying only one of these alleles seems to be sufficient for susceptibility to the disease [15]. These alleles can be situated in cis position (i.e. on the same chromosome): DR3-DQA1*05:01-DQB1*02:01 or in

trans position (i.e. on opposite chromosomes): DR5-DQA1*05:05-DQB1*03:01/DR7-DQA1*02:01-DQB1*02:02 [16, 17] (Figure 2). The risk for celiac disease is increased if having two copies of the DQB1*02 allele, a so-called gene dosage effect [18], which is coupled to an increased T cell response [19] and have also been associated to changes in the phenotype [20-22]. The majority of the remaining 5-10% of individuals with celiac disease carry the haplotype DR4-DQA1*03:01-DQB1*03:02 (DQ8) and a minority of these carry the haplotype DR7-DQA1*02:01-DQB1*02:02 (DQ2.2) [23] (Figure 2).

(19)

Figure 2. The HLA-haplotypes and DQ molecules associated with celiac disease. The HLA

heterodimer, consisting of the α-chain and the β-chain, is coded for by alleles within the DQA1-DQB1-region. The α-chain coded by DQA1*0501 or DQA1*0505 differ by only one amino acid in the leader peptide and similarly, the β-chain coded by DQB1*0201 or DQB1*0202 differ by one amino acid in the membrane-proximal domain [17, 24]. Modified from an original figure in Sollid, L.M. Annu Rev Immunol, 2000. 18: 53-81.

Not having any of the aforementioned HLA haplotypes is considered a rarity among celiac disease individuals where the majority of non-DQ2.5 and non-DQ8 individuals carry half of the DQ2 heterodimer [15]. Few specific HLA alleles have been suggested to be protective for the development of celiac disease [25], such as HLA-DQB1*06:02 in type 1 diabetes (T1D) [26], but recently HLA-DPB1*04:01 was found to down-modulate the risk for having tTGA in DR3-DQ2 positive children [27]. DQB1 DQA1 *02:01 CIS α-chain β-chain 3 Gluten *05:01 DQ2.5 DR HLA-region DQB1 DQA1 *02:02 TRANS α-chain β-chain 7 Gluten *05:05 *02:01 *03:01 5 DQ2.5 DR DQB1 DQA1 *02:02 α-chain β-chain 7 Gluten *02:01 DQ2.2 DR DQB1 DQA1 *03:02 α-chain β-chain 4 Gluten *03:01 DQ8 DR HLA-region

(20)

Overall, genes within the HLA region are estimated to account for around 40% of the genetic predisposition in celiac disease and recently several loci within this region but outside the HLA-DQ region were identified [28, 29]. Genome wide association studies of single nucleotide polymorphisms have revealed 39 non HLA loci that could be of importance [30, 31]. Most of these loci involve the immune response and are shared with other autoimmune diseases [32].

Environmental factors

In Caucasians, approximately 30-40% are either DQ2 or DQ8 carriers [33]. Since the majority of these are gluten consumers but only a minority acquire celiac disease there has to be additional factors contributing to the disease risk [34] (Figure 1). The rapid change in incidence of celiac disease in very young children between different birth cohorts in Sweden during the mid-1980-90s that coincided with change in infant feeding recommendations of gluten clearly indicated that environmental exposures are important for the disease development [35, 36]. However, to date no single causal factor has yet been identified although efforts have been undertaken to study the association between first exposure to gluten and breastfeeding during infancy and subsequent risk for celiac disease [37-40]. Other factors proposed to be associated with celiac disease are pre- and perinatal events [41-43], infections[44, 45], drug use [46], nutritional defects [47], dysbiosis [48] and food processing [49].

Immune response to gluten

There are four major factors interacting in the pathophysiological process of celiac disease: gluten, the enzyme tissue transglutaminase type 2 (tTG), the HLA-DQ2 or DQ8 heterodimer molecules, and T cells [50]. In celiac disease patients, gliadin peptides pass the epithelial barrier of the small intestinal mucosa, probably trough mechanisms involving increased permeability [51], and reach the lamina propria in which they are bound to HLA-DQ2 and/or DQ8 heterodimer molecules on APC and presented to reactive T cells [52]. Activated T cells elicit an inflammatory response involving the release of the intracellular tTG which extracellularly catalyses the amino acid glutamine to glutamic acid by deamidation and render the gliadin peptides more negatively charged [53]. This enzymatic reaction increases the binding affinity of the antigen to the cleft of the HLA-DQ2 or DQ8 heterodimers stimulating a stronger gluten specific CD4+ T cell response [54]. These gliadin-specific, HLA-DQ2 or DQ8 restricted T cells are only found in celiac disease patients and not in healthy individuals [55]. The activation of CD4+ T cells in turn stimulate the development of T helper cells type 1 (TH1), mediated by cytokines such as IFNγ, and the release of metalloproteinases by fibroblasts eliciting the typical histological injury in celiac disease characterized by increased

(21)

number of intraepithelial lymphocytes (IEL), crypt hyperplasia, and villus atrophy [56] (Figure 3).

Figure 3. The immune response to gluten in celiac disease (bottom picture) and the most common

histological mucosal changes (top picture). Clarifying block arrows added to original picture reproduced with permission from the publication of Green, P.H. et al. N Engl J Med, 2007. 357(17): 1731-43, Copyright Massachusetts Medical Society.

The role of the B cells in the pathophysiological process of celiac disease and its production of autoantibodies is still elusive [57]. The antibody production

Villous atrophy

Intraepithelial lymphocytosis Crypt hyperplasia

(22)

dependency of gluten intake in addition to strong association to HLA-DQ2, DQ8 or both, indicates the involvement of gluten-specific T cells in the activation of different B cell clones and subsequent production of antibodies against cross-linked deamidated gliadin and tTG [16]. Once B cells are activated they in turn are proposed to activate additional CD4+ T cells leading to a vicious circle of enhanced activation and subsequent chronic inflammation [50].

Parallel to the adaptive immune response in the lamina propria CD3+ T cells in the epithelial lining of the intestinal mucosa are activated. Gliadin seems to be directly toxic by inducing production of IL-15 and other cell-surface ligands in enterocytes in an innate immune response [58]. This, in turn, activates natural killer cells involved in the destruction of the intestinal epithelium leading to villous atrophy [59, 60].

Cytokines in celiac disease

Cytokines are proteins secreted by cells of the immune system and serve as mediators of various functions ranging from stimulating growth and differentiation of lymphocytes, activating effector cells and stimulating the development of hematopoietic cells. Their main effect is exerted locally (i.e. autocrine or paracrine action), but if secreted in sufficient amounts they can reach the circulation and act on sites distant from the site of production (i.e. endocrine action). Cytokines mediate effects both within the innate and adaptive immune response characterized by activation of T cells as well as being signature cytokines of different subsets of T helper cells, e.g. TH1 and TH2 [61].

Several cytokines seem to be involved in the immunologic process in celiac disease. IL-15, produced by epithelial cells and dendritic cells, induces epithelial apoptosis by affecting IEL [62] and is suggested to make intestinal T cells unresponsive to suppressive effects of regulatory T cells [63]. IL-15 also induces the production of IL-21 from T cells and IEL which co-express TH1 cytokines such as IFNγ [64]. IFNγ, the dominating cytokine produced by CD4+ T cells, induces the release of metalloproteinases by fibroblasts which degrade extracellular matrix and furthermore increase cytotoxicity of IEL [64]. IFNγ also enhances the HLA expression on APC thereby sustaining the response to gluten antigen [19]. Several studies have examined cytokine levels both locally in the intestinal mucosa [65, 66] as well as systemically in serum where they have been studied as markers of disease activity in affected individuals [67-71].

(23)

Prevalence

The prevalence of disease is defined as the proportion of a population having a certain condition e.g. a disease [72]. The prevalence can be measured at a certain time; point-prevalence, or during a period; period-prevalence. The prevalence varies depending on the population examined but also depending on method used for detecting affected individuals as well as definition of disease.

The prevalence of both screening-detected and clinically detected celiac disease in Sweden in 12-year old children has been estimated to between 2.2-2.9% depending on birth cohort studied [73]. In Norway, the prevalence of clinically detected celiac disease in 12-year old children or younger was estimated to 0.4% [74]. The prevalence of both clinically and screening-detected celiac disease in Finland in schoolchildren was 1% [75]. In contrast, the prevalence in Denmark is considered to be the lowest (0.08%) of the Scandinavian countries [76]. The prevalence of screening-detected and clinically detected celiac disease from other countries in Europe range from 0.3% [77] to 1.4% [78, 79] whereas prevalence in the United States is estimated to 0.8% [80]. In the non-western world celiac disease is believed to be as common as in Europe in the northern part of Africa, Middle-East and India, but is uncommon in China and adjacent countries [81]. All together, these prevalence studies have established the current belief that celiac disease is one of the most common chronic diseases both in adults and in children. To date, there is no plausible explanation to why Sweden has one of the highest reported prevalence numbers of celiac disease in the world.

Celiac disease is also known to be more prevalent in certain disease risk groups [82]. The largest risk group is first and second degree relatives of celiac disease individuals (prevalence of celiac disease 10-20%) [80]. Having another autoimmune disease, such as T1D (3-12%) [83], autoimmune thyroiditis (up to 7%) [84] or autoimmune liver disease (12-13%) [85], confers an increased risk to develop celiac disease compared to the general population. Also, having IgA deficiency 8%) [86], Downs’s syndrome (5-12%) [87], Turner syndrome (2-5%) [88] and William’s syndrome (up to 9%) [89] are linked to the comorbidity of celiac disease. This increased risk has led to the recommendation of routine screening among these groups of patients in the clinical setting [82].

In individuals carrying the HLA-DQ risk alleles, longitudinal prospective population studies that follow at genetic risk birth cohorts selected from the general population have clearly shown that the incidence of celiac disease varies depending on HLA genotype. The DAISY study (Diabetes and Autoimmunity Study in the Young) from Denver in Colorado, United States, showed that the cumulative incidence of celiac disease in children carrying HLA-DR3-DQ2/X (i.e. heterozygotes for DQ2) was 3.4% by 5 years of age [90]. In the DIPP study

(24)

(Diabetes Prediction and Prevention) from Finland, following children with HLA-DQB1*02, the prevalence of celiac disease was 1.3% by mean 5 years of age [91]. The multicentre study TEDDY (The Environmental Determinants of Diabetes in the Youth), involving six clinical centres in four different countries, has revealed that the incidence of celiac disease is dependent on HLA-genotype in young children [22]. However, none of these prospective studies have enrolled children having HLA alleles not associated to celiac disease and thus the prevalence of celiac disease in this HLA-non-risk group is not known. The prevalence of celiac disease thus varies substantially between geographical regions and different subpopulations, which emphasises that both genetic and environmental factors are important for its widespread occurrence. More importantly, the reported prevalence also seems to be dependent on whether screening of the general population has been performed or not highlighting the importance of method for identification of disease.

Clinical Presentation

Celiac disease is associated to a myriad of symptoms and manifestations, which range from classical gastrointestinal symptoms including malabsorption to more subtle variants. The description of the disease has changed over time, possibly because of increased knowledge about disease manifestations, but probably also because of true changes in the clinical pattern towards milder forms and more often extra-intestinal symptoms and older age at diagnosis [36, 92-94]. Classically, in young children (< 3 years), symptoms often include intestinal symptoms like diarrhoea and abdominal distension but also extra-intestinal symptoms like growth retardation. Older children (<18 years) often present with constipation, abdominal pain and extra-intestinal manifestations such as fatigue, delayed puberty, and arthralgia [95]. The extra-intestinal symptoms like iron deficiency anaemia, elevated liver enzymes, skin lesions of dermatitis herpetiformis, neurological manifestations such as neuropathy, and psychiatric disorders like depression needs knowledgeable medical personnel to be identified as symptoms of celiac disease [96].

More importantly, many celiac disease patients can be virtually asymptomatic (i.e. lack of symptoms even in response to direct questioning at diagnosis), or have

subclinical forms (i.e. having insufficient symptoms to raise the suspicion of

celiac disease in clinical care) [97]. In fact, prospective birth cohort studies have revealed that the majority of children with celiac disease that are identified by screening only exhibit mild symptoms [98, 99] and would most certainly be unrecognized without the screening procedure [100]. Thus, the understanding of

(25)

celiac disease presentation to date points in the direction of symptoms not being reliable tools for identifying disease and other means of recognition is warranted.

Diagnosis

The diagnostic work-up for celiac disease of today constitutes of a combined arsenal of disease specific serological markers, histology and genetic tests.

Serology

Since the 1980s, the first step towards a celiac disease diagnosis has been the analysis of presence of autoantibodies in serum and today, one of the most widespread applied antibody tests is the detection of IgA autoantibodies directed against tissue transglutaminase (tTGA) [3]. Different methods for detection have yielded high sensitivity between 90-95% and specificity around 99-100% [101]. Autoantibodies are analysed using either solid phase enzyme linked immunosorbent assays (ELISA) or liquid phase radioligand binding assays (RBA) and the two methods have shown a good correlation [102].

Autoantibodies directed against the endomysium, IgA-EMA, were discovered before tTGA and have also yielded high sensitivity and specificity above 90% for celiac disease [103]. IgA-EMA is analysed by indirect immunofluorescence, which is considered expensive and associated with inter observer variability not found in ELISA or RBA.

IgG antibodies against deamidated gliadin peptides (DGP), developed the last 15 years, have comparable sensitivity and specificity as IgA-tTGA [103] and in the revised ESPGHAN diagnostic criteria for celiac disease, IgG-DGP is recommended to be used in patients with IgA deficiency and in children younger than 2 years of age [82]. As both IgA-tTG and IgA-EMA display a lower sensitivity in children younger than 2 years of age compared to older children, IgA anti-gliadin antibodies (AGA) could be an option, but should be interpreted with caution because of its low specificity [104] and in a recent study IgA-tTGA showed sufficient sensitivity even in very young children [105]. Since celiac disease is more common in IgA deficient individuals, it is important that all IgA based tests are accompanied by measurements of IgA levels in serum [106]. In IgA deficient celiac disease individuals, IgG-tTGA and IgG-DGP have shown comparable sensitivities as for IgA-tTG in IgA-sufficient cases [107].

Finally, gluten consumption is a prerequisite for reliability of detection of all of these aforementioned antibodies thus the tests can also be used as markers for response to treatment and dietary compliance.

(26)

Histology

The assessment of duodenal biopsies has long been considered the gold standard of celiac disease diagnostics. During the past decades, the perception of celiac disease being an autoimmune disease not just affecting the intestinal mucosa has questioned this belief. The discovery of disease specific autoantibodies and the development of reliable methods have led to the recommendation that histological diagnosis can be omitted among children in clear cut cases with high autoantibody levels [82]. However, in unclear cases with low levels of tTGA and subtle symptoms or no clinical signs of the disease, intestinal biopsies taken via upper endoscopy or via capsule still offers an important diagnostic approach. On the contrary, histological diagnosis is still recommended for all adults due to e.g. differential diagnoses such as malignancies or refractory disease [108]. It should be kept in mind that the affected intestinal mucosa in celiac disease patients may be patchy [109], why it is recommended to take multiple biopsies from the duodenum as well as from the bulb [110].

The typical microscopic histological alterations in celiac disease include infiltration of plasma cells and lymphocytes in the lamina propria, increased numbers of IEL, elongated crypts, and partial or total villous atrophy [111] (Figure 3). These histological findings are classified according to the Marsh-Oberhuber classification in which the normal mucosa is denoted Marsh 0. The infiltrative lesion is defined as a normal mucosal architecture but increased number of IEL; > 40 IEL/100 enterocytes called Marsh 1. In the hyperplastic lesion there is still a normal villous architecture but an increased number of IEL in combination with crypt hyperplasia which is called Marsh 2. In the next stage, called the destructive lesion, increased IEL, crypt hyperplasia and various degrees of villous atrophy are present (partial villous atrophy- Marsh 3A, subtotal villous atrophy- Marsh 3B and total villous atrophy- Marsh 3C). The hypoplastic lesion, called Marsh 4, is defined as a flat mucosa with normal count of IEL and normal crypts [112, 113]. It is important to notice that the histological alterations are not pathognomonic of celiac disease and can be found in other conditions such as cow’s milk protein allergy, giardiasis, Crohn’s disease, inherited enteropathies, and drug reactions [111]. Additional diagnostic histological markers are therefore being developed such as count of IEL at the villous tip, γδ+ IEL, and intestinal deposits of tTGA [114, 115]. Finally, the inter-observer variability is important to consider when histological specimens are being interpreted.

Genetic testing

The strong negative predictive value of HLA-DQ2, DQ8, or both, for celiac disease makes genotyping an applicable method for selecting individuals at genetic risk. HLA genotyping can be assessed by polymerase chain reaction (PCR)

(27)

and hybridization with allele specific probes; a highly specific but time consuming and expensive method. However, the positive predictive value is low as 30-40% of populations in the Western World (Caucasian, White or European origin) are carriers of HLA-DQ2 and/or DQ8 haplotypes [33, 116] but only about 1% develop celiac disease. Thus, the clinical value of testing for HLA-DQ risk alleles has mostly been in the case of excluding celiac disease in dubious cases [117]. Alternative methods for detection of HLA-risk alleles in celiac disease such as single nucleotide polymorphisms (SNPs) are currently being developed [118, 119], but so far testing for non-HLA genetic risk loci has only been done in the research setting [120].

Diagnostic criteria in symptomatic celiac disease

The first diagnostic criteria for celiac disease, established by the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) in 1969, recommended the use of intestinal biopsies as the gold standard of methods to confirm villous atrophy at diagnosis, mucosal healing after gluten-free diet and relapse of the disease after gluten challenge. After several reviews, the ESPGHAN criteria were revised in 1990 to include only one initial diagnostic biopsy and thereafter a clear cut response to gluten-free diet with normalisation of serological markers as combined criteria [121]. With increased experience of the high diagnostic performance of tTGA and EMA, ESPGHAN revised the guidelines in 2012 and accepted a tTGA level > x10 the upper limit of normal confirmed by EMA and HLA as a replacement for the biopsy as the gold standard [82]. Still, an intestinal biopsy is recommended in dubious cases and tTGA levels should seroconvert to normal on a gluten-free diet [82].

Diagnostics in screening-detected celiac disease

In asymptomatic children at genetic risk, tTGAs are considered reliable tools for identifying celiac disease [122], but low levels of tTGA are more common than in symptomatic children [123]. The ESPGHAN guidelines therefore recommend HLA genotyping as the first step followed by tTGA as a second test in screening individuals at increased risk for celiac disease, such as first-degree relatives of celiac disease individuals, T1D patients, individuals with IgA deficiency, and Down syndrome [82]. In first-degree relatives and children with T1D there is a high rate of normalisation of tTGA despite a gluten containing diet [124, 125]. In addition, asymptomatic children detected with elevated levels of tTGA are recommended a duodenal biopsy to confirm diagnosis [82].

Follow-up of childhood celiac disease should include evaluation of symptoms and check-up of tTGA to control response to treatment and dietary compliance, but a control biopsy is not routinely recommended [82]. The proportion of mucosal

(28)

recovery in children is considered to be high, but there is up to date limited evidence in children to support any of the applied follow-up measurements although serology seems to have a high negative predictive value [126].

Screening

Medical screening consists of three elements. First, it is a process of selection with the purpose of identifying those individuals who are at a sufficiently high risk of a specific disorder to warrant further investigation or sometimes direct preventive action. Second, screening should be systematically offered to a population who would not seek medical attention on account of symptoms of the disease for which screening is being conducted. Third, the purpose of identifying a disease should benefit the individual [127].

In the 1960s, Wilson and Jungner published for the World Health Organization (WHO) ten criteria to be fulfilled for a disease to be considered for mass screening [128]:

1. The condition sought should be an important health problem.

2. There should be an accepted treatment for patients with recognized disease.

3. Facilities for diagnosis and treatment should be available.

4. There should be a recognizable latent or early symptomatic stage. 5. There should be a suitable test or examination.

6. The test should be acceptable to the population.

7. The natural history of the condition, including development from latent to declared disease, should be adequately understood.

8. There should be an agreed policy on whom to treat as patients.

9. The cost of case-finding (including diagnosis and treatment of patients diagnosed) should be economically balanced in relation to possible expenditure on medical care as a whole.

10. Case-finding should be a continuing process and not a “once and for all” project.

These criteria have been modified since then by different health organisations in various countries but they still remain the basis when mass screening is considered.

(29)

Screening is a method of prevention, i.e. any activity that reduces the burden of mortality or morbidity from disease. Primary prevention is aimed at preventing the development of disease, in celiac disease by inducing tolerance to gluten, which to date is not possible, but much aim is focused on if early infant feeding could be optimized and intervention by vaccination in risk individuals [129]. On the contrary, screening for celiac disease could be beneficial at a secondary as well as at a tertiary prevention level. As secondary prevention, identification of individuals with unrecognized disease, possibly at an early stage, increases the opportunities for interventions to halt disease progression and the emergence of symptoms. As tertiary prevention, identifying undiagnosed individuals and instituting a gluten-free diet will reduce the negative impact of an already existing disease by restoring function and reducing disease-related complications.

Mass screening for celiac disease

Most of the screening criteria are fulfilled for symptomatic celiac disease, whereas in subclinical and asymptomatic celiac disease the benefits of diagnosis and treatment are still a matter of debate [129-132].

First, it has not yet been fully clarified whether having asymptomatic celiac disease confers a health problem to the individual (i.e. criteria 1). An argument for early identification is the finding of low bone mineral density in young children with subclinical or potential celiac disease [133]. In addition, young children found in screenings often display symptoms albeit they are mild [98, 99]. On the other hand, health-related quality of life (HRQoL) prior to diagnosis seems not to be affected [134].

Second, performing an intestinal biopsy during anaesthesia in tTGA positive individuals is an invasive procedure and expensive method associated with risks, which could be questioned in asymptomatic children (i.e. criteria 6), although a recent study has demonstrated that affected children seems to be positive towards screening in general [135].

Third, there is a lack of information on the health benefits of treatment in screening-detected celiac disease and potential negative effects of a restricted diet and reduced compliance (i.e. criteria 8). HRQoL, symptom reduction and compliance after starting a gluten-free diet show overall positive results in children detected by screening [136-140] as well as in adults [141]. Although studies on mortality have not yet examined screening-detected celiac disease in children [142, 143], the mortality in adults with celiac disease detected by screening seemed not to be affected [144].

Fourth, screening will probably identify a number of individuals with celiac disease autoimmunity (i.e. positive tTGA) or potential celiac disease (i.e. tTGA positive with normal mucosal features). The long-term risk of complications in

(30)

children with celiac disease autoimmunity with no or only mild histological alterations have not been fully explored (i.e. criteria 7 and 8). Several studies point in the direction that a large fraction of children with celiac disease autoimmunity and potential celiac disease lose their tTGA or do not develop mucosal changes at follow-up and therefore screening could create unnecessary concern [91, 125, 145, 146]. Mild histological alterations have so far not been considered diagnostic for symptomatic celiac disease but there are studies on both children and adults in favour of treatment [143, 147-149].

Finally, there is a lack of cost benefit studies of mass screening in children and how a screening procedure could be implemented in the general health care system still needs to be established (i.e. criteria 9) [150, 151].

Treatment

To date, the only recommended treatment for celiac disease is a strict gluten-free diet, i.e. a diet free from wheat, rye and barley, which is considered a safe and harmless treatment with nutritional and energy contents comparable to a normal diet [152]. Non-adherence is coupled to certain risk-factors such as young age at diagnosis but does not appear to be related to symptoms at diagnosis [153] and compliance in patients with celiac disease detected by screening seems to be high [137, 138]. The social burden of a gluten-free diet should not be overlooked, especially during adolescence when dietary advice should be modified accordingly. Based on this assumption, alternative therapies are being developed. Glutenases that enzymatically degrade gluten to non-toxic peptides, polymers that bind gluten intra-luminally, tight junction modulators such as zonulin-antagonists, and blocking of lymphocyte homing in the intestines, are potential drugs currently under development [154].

Associated diseases and long-term complications

Celiac disease is associated with other diseases and long-term complications later in adulthood if left untreated. Hitherto, the associated risks have been assessed in studies of individuals with clinically diagnosed celiac disease whereas the risks in individuals with subclinical and screening-detected disease are unresolved.

The overall risk for an individual with celiac disease to develop another autoimmune disease is increased affecting 15-20% of celiac disease patients [155] of which thyroiditis is the most prevalent (3-10%) [156]. In addition, the risk for T1D in celiac disease individuals is more than twice that of the general population [157] and the prevalence of celiac disease in T1D patients is approximately 6%

(31)

[83]. The reasons for the co-occurrence of autoimmune diseases is unknown but is probably mainly due to the shared HLA haplotypes and possibly other identified risk genes outside the HLA region [158]. However, it has been speculated whether gluten may be a trigger also for T1D, either as a triggering antigen on the pancreatic islet beta cells or indirectly by causing an increased intestinal permeability for other antigens such as enteroviruses [159], but whether the risk of other autoimmune diseases in celiac disease is modulated by a gluten-free diet remains unclear [160, 161].

The probably most feared of complications is the risk for malignancies, which have been extensively studied in patient register studies [162-164]. Celiac disease individuals have a slight increased risk for malignancies, preferentially during the first year after diagnosis after which the risk levels off comparable to the general population [165]. Malignancies associated with celiac disease are non-Hodgkin lymphoma, especially enteropathy-associated T-cell lymphoma (EATL), which in turn is closely linked to refractory celiac disease type 2, and gastrointestinal adenocarcinoma [166]. However, these are rare diseases preferably presented in adults and the absolute risks in celiac disease individuals are low [166]. Also, mortality risk is increased in both adults as well as children with celiac disease, but decreases by time after diagnosis [143].

Other long-term potentially severe complications are pregnancy related complications which have been found to be more common in untreated women with celiac disease in contrast to treated individuals in some [167], albeit not in all studies [168, 169]. Infertility has been discussed as a long-term complication of celiac disease, but register based studies from Sweden do not support this [170, 171].

In contrast to the aforementioned complications, the risk of osteopenia or osteoporosis and related fractures in celiac disease is far more common [172, 173]. In a study on middle-aged women selected from the general population, high levels of tTGA were associated with lower bone mineral density (BMD) and higher fracture frequency in women between 50 and 64 years of age [174]. More importantly, a gluten-free diet improves BMD in adults within a year although not all individuals normalize their bone mass at long-term follow-up [175, 176]. The risk of low BMD and osteoporosis have also been studied in children and adolescents with celiac disease, but there are some circumstances that needs to be considered when it comes to the study of bone mass during childhood.

Bone mass during childhood

The skeleton accumulates bone mineral up till the end of the second or the beginning of the third decade of life to reach its peak bone mass (PBM); the highest bone mass value in life [177]. Once the PBM reaches its plateau the bone

(32)

mass starts to decline which accelerates with increasing age, particularly in women during the 10 year period after menopause [178]. The most rapid accrual of bone mineral occurs in girls between 11-14 years of age and in boys between 13-17 years of age, respectively [179]. PBM is affected by different life-style factors; physical activity, nutrition, chronic diseases, but the most important regulator are hereditary factors [177, 180] (Figure 4).

Figure 4. Bone mass acquisition and changes during life and factors affecting outcome. Reprinted

from Heaney et al., Osteoporos Int, 2000. 11(12): 985-1009 with permission from the publisher.

PBM is an important predictor of osteoporosis later in life and adult decrease in BMD increases risk for fractures [181, 182]. Children affected by a chronic disease during growth, such as inflammatory bowel disease, are at an increased risk to gain a lower PBM thus affecting BMD both during childhood and adult life [183]. However, it is important to underline that a low BMD does not solely cause fractures and that fracture risk is affected by several other factors that can coincide with low BMD, such as low muscle strength. It is also important to know that bone tissue is under constant turnover throughout life, a process called remodelling [184].

Bone mass measurement

Bone mass is often expressed as bone mineral content (BMC) (g), which refers to the mineral detected at measurement or areal BMD (g/cm2),which is bone mineral detected over a projected area. There are several methods to assess BMC and BMD of which dual X-ray absorptiometry (DXA) is considered the gold standard for bone measurement [185]. This technique is based on filtered X-rays of two

(33)

energies, which enables the distinction between soft tissue and bone, and a detector on the other side of the measured individual that can be used to calculate bone mass. Compared to other methods it can scan all parts of the body, it is a fairly rapid procedure (a total body scan of a 10-year old child takes approximately 7 minutes), it produces a low radiation dose (1-8 µSv per scan corresponding to 1/1000 of the yearly background radiation) and combines a high accuracy (the closeness of a measurement to the true value) as well as a high precision (the reproducibility of a measurement when repeated) [186].

Osteoporosis in children is defined as the presence of a clinically significant fracture history and low BMC or BMD. In children, a low BMC/BMD, defined as a Z-score (number of SD from the mean of a population adjusted for age and gender) < -2.0, should be used instead of the term osteopenia [187]. If osteoporosis is related to another disease or medication it is often referred to as secondary osteoporosis [188].

BMD in children with celiac disease

Children with celiac disease have reduced BMC and BMD at diagnosis compared to healthy individuals [189, 190]. A gluten-free diet increases BMD to normal levels within a year of treatment in some previous studies [191-194], and BMD is maintained on a long-term follow-up [195, 196], whereas other studies have shown lack of normalisation after 1 year [197] and 2 years of treatment, respectively [198]. There is also some evidence that children who are older at diagnosis have lower BMD adjusted for age compared to younger children indicating that age at diagnosis influence outcome of BMD [199-203]. The dietary compliance has also been evaluated in favour of a strict gluten-free diet [204, 205]. Neither an association between presence of clinical symptoms and BMD have been found [201, 206], nor between BMD and levels of tTGA, histological alterations or laboratory measurements such as parathyroid hormone (PTH) [198, 206].

Studies on biomarkers of bone-formation and bone-resorption have shown an increased activity of resorption markers and decreased activity in bone-formation markers at diagnosis, which is reversed after start of a gluten-free diet [194, 207]. The majority of studies do not demonstrate changes in vitamin D metabolites or in PTH levels in children with untreated celiac disease [195, 200, 208]. Fracture risk in children with celiac disease is not well-studied, although the absolute risk of associated hip fractures in childhood celiac disease seems to be low [209].

However, many of these previous studies are performed on clinically detected celiac disease. In adults, studies on patients with celiac diseases detected by screening [210] as well as asymptomatic individuals with a former celiac disease

(34)

diagnosis but on a gluten containing diet [211, 212], all demonstrate the presence of osteopenia/osteoporosis in the majority of patients. Only one prospective cohort study of screening-detected tTGA positive asymptomatic preschool children have been performed showing a lower BMD in screening-detected subjects compared to tTGA negative controls [133]. This highlights the need for additional studies of BMD in children identified in screenings.

Causes of low BMD in childhood celiac disease

There are several possible factors that can interact to alter bone metabolism in celiac disease but there is up to date no clear evidence for a specific main cause [213] (Figure 5).

Figure 5. Possible mechanisms involved in bone loss in celiac disease.

Malabsorption of calcium and vitamin D in the duodenum and jejunum and secondary hyperparathyroidism is one possible cause for increased bone resorption and decreased formation shown in adults in contrast to children [200, 214]. Theoretically, malnutrition of these nutrients as a result of secondary lactose intolerance found in celiac disease could be another possibility, but is probably of

INTESTINAL INFLAMMATION CALCIUM MALABSORPTION VITAMIN D MALABSORPTION HYPOCALCAEMIA SECONDARY HYPERPARATHYROIDISM ↓ BONE MASS ↑ OSTEOCLAST ACTIVITY ↑ BONE RESORPTION ↑ IL-1, IL-6, TNFα ↓ OPG/RANKL ratio Antibodies directed against bone tissue

↓ intestinal absorption of calcium ↓ kidney reabsorption of calcium

(35)

less importance in countries where lactose-free products are calcium and vitamin D enriched and easily accessible. Severe disease with malnutrition and fatigue leading to reduced physical activity in combination with reduced sun exposure resulting in vitamin D deficiency could in theory contribute to low BMD.

Inflammation derived cytokines have long been known to alter bone formation, leading to bone loss in chronic paediatric inflammatory conditions such as in inflammatory bowel disease [215], rheumatologic diseases [216] as well as in celiac disease. The cytokines IL-1β, IL-6 and TNF-α are able to activate osteoclasts (i.e. bone cells in charge of bone resorption) and have been found elevated in adults with untreated celiac disease [217]. In synergy, IL-12 and IL-18 associated to inhibition of osteoclasts have been found to be down-regulated in adult celiac disease individuals [218]. Most cytokines are thought to regulate bone remodelling directly or indirectly through the RANK/RANKL/OPG pathway [219]. Receptor activator nuclear factor κB (RANK) is located on pre-osteoclasts. When receptor activator nuclear factor κB ligand (RANKL) is bound to RANK this stimulates the differentiation and activation of osteoclasts, which promotes bone resorption. Osteoprotegrin (OPG) acts as a decoy receptor for RANKL and inhibits RANK/RANKL interaction and therefore inhibits osteoclast activity. Both RANKL and OPG are secreted by osteoblasts (i.e. bone cells responsible for bone formation) as well as by other tissue cells. The OPG/RANKL ratio has been found to be lower in celiac disease patients compared to healthy individuals, implicating its role in bone loss in celiac disease [218, 220].

Autoantibodies produced as a part of the inflammatory process in celiac disease have also been postulated to play a role in bone loss. Sera from untreated celiac disease individuals provide evidence of presence of IgA autoantibodies directed against bone specific tTG, which is an enzyme important for bone matrix stabilisation and mineralisation [221]. Autoantibodies directed against OPG have been found in celiac disease patients associated with lower BMD [222, 223], but the clinical significance of testing for these antibodies are still contradicting [224]. Despite the body of evidence that BMD is affected also in children with celiac disease, there is no consensus whether DXA should be performed as a standard procedure on a regular basis. Furthermore, there is no other treatment except for gluten-free diet recommended in children with celiac disease [225].

(36)

Aims

The overall aim of the research and the basis for this thesis was to identify children with screening-detected celiac disease in an HLA-genotyped birth cohort and to study subclinical manifestations of the disease.

The specific aims were:

I. To estimate the prevalence of screening-detected celiac disease in an

HLA-genotyped cohort of children at 3 years of age (Paper I).

II. To estimate the prevalence of screening-detected celiac disease in a follow-up study of a cohort of HLA-genotyped children at 9 years of age (Paper II).

III. To examine signs of systemic inflammation by measuring cytokines in

serum of 3-year old children with screening-detected celiac disease at time of diagnosis and after treatment with a gluten-free diet (Paper III).

IV. To assess if 9-year-old children with screening-detected celiac disease have affected bone mineral density (BMD) and/or levels of 25(OH) vitamin D3 and parathyroid hormone (PTH) at time of diagnosis and under treatment with a gluten-free diet (Paper IV).

(37)

(38)

Methods

Population and study design

This thesis is based on the Celiac Disease Prediction in Skåne (CiPiS) study; a prospective population-based cohort study aimed at identifying celiac disease by screening in children born in the region of Skåne 2001-2004. Skåne is a province situated in the southern part of Sweden with 1.2 million inhabitants (2004) in which approximately 12.000 children were born yearly during the time period 2001-2004 (www.scb.se) (Figure 6).

Figure 6. The Celiac Disease Prediction in Skåne (CiPiS) study is based on a cohort of children born

in the region of Skåne, Sweden 2001-2004.

DiPiS study

The CiPiS study is part of the DiPiS (Diabetes Prediction in Skåne) study which is a prospective cohort study aimed at determining the predictive value of genetic risk in combination with diabetes autoantibodies and to identify non-genetic risk factors for T1D in all children born in Skåne during the years 2000-2004 [226]. During the study period, all parents of newborn children were invited to participate in the DiPiS study at delivery, and after oral consent, cord blood was collected for HLA genotyping, analysis of islet autoantibodies, and a registration form was filled in. The form contained information about the mother’s age, date of birth, multiple births, gestational age, gender, diabetes in mother, gestational diabetes, and maternity clinic. When the child was two months of age, parents were sent a

(39)

letter of invitation to continue participation and participating parents gave their written informed consent and filled in questionnaires concerning hereditability for diabetes and factors relevant to diabetes during pregnancy, delivery and the first two months of the child’s life [226, 227]. Parents of participating children who had increased risk for developing T1D were contacted again when the child was two years of age for continued participation and a blood sample was drawn from the child and sent in for analysis of islet autoantibodies and thereafter annually. The increased risk for T1D was mainly based on the HLA genotype of the child, but other risk factors was also taken into account (heredity for diabetes, infections during pregnancy, mothers age >40 years, islet autoantibodies in cord blood, large for gestational age and postnatal jaundice).

CiPiS study

All children born between June 2001 and August 2004 whose parents answered the two-month questionnaire in DiPiS were eligible for participation in the CiPiS study. According to Statistics Sweden (www.scb.se) approximately 39,000 children were born in Skåne during this period. A DiPiS-registration form at delivery was received from 29,913 newborns and no information was obtained from the approximately 9000 remaining newborn children. Successful HLA genotyping was available from 29,039 children and parents of 19,621 children gave written consent to continue participation at two months (Figure 7). All children having either HLA-DQB1*02 or *03:02, or both alleles, were considered having genetic risk for celiac disease. Children having HLA-DQA1*05:01, also conferring risk for celiac disease, were not all identified since only children having HLA-DQB1*02 were HLA-DQA1-genotyped (see paragraph about HLA-DQ genotyping) [15, 228]. The HLA-DQB1 distribution of the genotyped population, of children to parents accepting participation in DiPiS study (answering the questionnaire), and of children invited to CiPiS study is shown in Table 1. During the study period, the HLA-DQB1*02 and DQB1*03:02 alleles occurred in 56.8% of the HLA-genotyped population comprising of more than 70% of the children born in the studied region.

At 3 years of age, two groups were invited to participate in the CiPiS study. Children having either HLA-DQB1*02, DQB1*03:02, or both alleles, were included in the HLA–risk group. Children not carrying any of these risk alleles were included in the HLA-non-risk group. The HLA-risk group were recruited mainly from the DiPiS invited group constituting of 90% of children having the high risk HLA-alleles for T1D and 40% of children having neutral risk for T1D (Table 1). In addition, 15% of the children carrying a protective allele for T1D were also invited to the CiPiS study (Table 1). Children without any risk alleles were invited regardless of if they were invited to the DiPiS study. In all, 13,860

(40)

children were invited at 3 years of age of whom 6206 children belonged to the HLA-risk group and 7654 children to HLA-non-risk group (Figure 7).

At 9 years of age, the same group of children were re-invited to a follow-up screening (Table 1). All children who had been diagnosed with screening-detected celiac disease at 3 years of age, those who had moved, died or denied continued participation, were excluded. Altogether, 13,024 children were invited in the follow-up screening at nine years of age of whom 5947 belonged to the HLA-risk group and 7077 to the HLA-non-risk group (Figure 7).

Figure 7. Children accepting participation in DiPiS (Diabetes Prediction in Skåne) study and

children invited to CiPiS (Celiac Disease Prediction in Skåne) study. *HLA risk in CiPiS is defined as having HLA-DQB1*02 and/or *03:02 alleles. ** Controls are HLA-non-risk group defined as children not having HLA-DQB1*02 and/or *03:02 alleles.

Parents of invited children were sent an information letter and were asked to respond regardless of they wanted to participate or not. They were also asked to respond if the child already had celiac disease and these children were excluded from the screening. In the screening at 9 years of age, non-responding parents were sent a letter of reminder and if still not responding they were reminded by phone call from a study nurse. Parents (one parent in the first screening and, if possible, both parents in the follow-up) of participating children gave their written consent and if so, they were sent a tube for collection of a blood sample at their local health care centre. In the screening at 9 years, they were also offered to have

Invited to CiPiS study at 9 years of age Invited to CiPiS study at 3

years of age HLA risk in CiPiS study

Children accepting participation in DiPiS study

at 2 month of age Children HLA-genotyped

in DiPiS study Children born in Skåne

June 2001- August 2004 39,000 29, 039 19,621 HLA-risk* 11,359 6206 5947 Controls** 8262 7654 7077 No respons No to participation in DiPiS Missing cord blood No to participation in DiPiS Moved, died Not invited in DiPiS follow-up No to participation in DiPiS Moved, died Celiac disease at 3 years

Searching for Celiac Disease Screening-detected celiac disease in an HLA-genotyped birth cohort

Searching for Celiac Disease Screening-detected celiac disease in an HLA-genotyped

birth cohort

Björck, Sara

Searching for Celiac Disease

Screening-detected celiac disease in an

HLA-genotyped birth cohort

Sara Björck

Searching for Celiac Disease

Screening-detected celiac disease in an

HLA-genotyped birth cohort

To all children having celiac disease;

known and unknown.

Table of Contents

Abbreviations

List of papers

Abstract

Background

Introduction

Pathogenesis

Gluten

Genetics

Environmental factors

Immune response to gluten

Cytokines in celiac disease

Prevalence

Clinical Presentation

Diagnosis

Serology

Histology

Genetic testing

Diagnostic criteria in symptomatic celiac disease

Diagnostics in screening-detected celiac disease

Screening

Mass screening for celiac disease

Treatment

Associated diseases and long-term complications

Bone mass during childhood

Bone mass measurement

BMD in children with celiac disease

Causes of low BMD in childhood celiac disease

Aims

Methods

Population and study design

DiPiS study

CiPiS study